A multidisciplinary team will establish a population laboratory for studying the management of depression in a major primary care clinic (23,000 adult patients) of a large HMO. Within this population laboratory, and with comparison physicians and patients drawn from other clinics of the HMO, the research team will develop and evaluate methods of treating depression well suited for wide-spread use in large primary care clinics. This research will evaluate a comprehensive program for primary care of depression targeting physician and patients skills in four areas: (1) recognition of depression and its effects; (2) monitoring depression to improve treatment and prevent relapse; (3) improving physician management of and patient adherence to antidepressant medication regimens; and (4) improving problem solving skills of patients. STUDY 1, Evaluation of physician interventions: Within the intervention clinic, all 16 physicians will be trained in the above methods of treating depression. A comparison sample of 16 physicians will be identified from other HMO clinics matched on tricyclic antidepressant (TCA) prescribing patterns. Subsequent to training, a cohort of patients receiving a new TCA prescription for depression will be identified for the intervention physicians (N=250) and for the control physicians (N=250). TCA therapy will be tracked using automated pharmacy data for these patients for one year. A random sample of 125 of these patients (per group) will be interviewed at baseline and at 3, 6 and 12 months after their index visit. These interviews will assess: depression diagnostic status; patient perceptions of physician expertise in treating depression, patient- physician collaboration, MD enhancement of patient control over depression; adherence to TCA medication regimens; TCA side-effects; symptomatic outcomes (depression, anxiety, somatization, chronic pain and fatigue) and behavioral outcomes (disability and health services utilization). STUDY 2, Evaluation of patient interventions: Screened intervention clinic patients (N=9750) with a positive depression screen (8%, N=780), referred by their physician for educational classes (65%, N=505), who meet definite or probable criteria for DSM-III-R major depression (66%, N=335) and who accept randomization (75%, N=250) will be randomly assigned to either: (1) educational case management provided by a psychiatrically skilled nurse (N=125); or (2) usual care by the patient's primary care physician (N=125). The nurse case manager will teach patients specific skills in pharmacologic and behavioral management of depression. Randomized patients will be interviewed at baseline and at 3, 6 and 12 months post-randomization. These interviews will assess the same process and outcome variables as Study 1.